Haskell library to robustly parse unstructured HTML

 TagSoup extracts information out of unstructured HTML code, sometimes
 known as tag soup. TagSoup does not require well-formed or
 standards-compliant HTML, or HTML that renders correctly in any
 particular rendering engine. TagSoup transforms HTML into a list of
 open tags with attributes, close tags, and text, but makes no attempt
 to group these together into any kind of structure.
 .
 This package is not related to the Java TagSoup library.